There are 2.5 million children living the HIV/AIDS worldwide. Most live in resource- limited settings. Treatment guidelines for children are often based on findings in adults and seldom assembled from studies employing a randomized design. This rationale led the NIH to initiate the PREDICT study underway in Thailand and Cambodia. PREDICT uses a randomized design to determine the optimal timing of antiretroviral initiation in children. International guidelines for ART initiation in children older than 12 months continue to recommend deferring therapy until symptomatic disease or immune compromise occurs, despite studies in younger infants (CHER trial) identifying a benefit for early treatment. Treatment outcomes for older children are needed and neurodevelopmental outcomes could critically inform guidelines. This proposal will add robust neurodevelopmental and imaging outcomes to PREDICT. This work will extend our current limited evaluation of neurodevelopmental outcomes to determine if differences in long-term neuropsychological and neurological outcomes exist in relation to deferring antiretroviral therapy. We will also enroll needed HIV-negative comparative groups of HIV-exposed and unexposed children. Since microencephaly is a common manifestation of HIV in children, volumetric analyses may be most informative. Should the primary findings from PREDICT fail to identify differences by randomized group (immediate compared to deferred therapy), findings from the proposed work may have an enormous impact by informing treatment guidelines worldwide.